How to learn 'finance', investing and trading from scratch

Thanks to my background(CA student, CS/CWA ranker) and being the lone commerce student among a bunch of engineers, I'm often asked questions by juniors and peers alike about how to understand the aforementioned subjects. While I do not claim expertise in any of them-indeed I think the day someone stops learning is when he does that-I've read/invested/traded/learnt/done a reasonable amount to identify what works for a learner-and more importantly what does NOT work. So without much ado, I present below some points for someone from a non commerce background, to approach the field of finance. As many of the points would show, I'm  a great believer in the value of self education.

1. Finance is not one huge homogenous entity: Take some time to appreciate the huge variety from asset management to corporate finance to investment banking. Whether you like finance, marketing, operations or HR, chances are that finance needs someone like you.
2. Get familiar with the jargon by reading good financial papers:-And no, I no longer consider Economic Times a good paper, due to the overdose of commercial content and tendency to confuse reporting with merely reproducing corporate press releases. While their supplements are good on Sundays/Mondays, that is the best that can be said. Instead, I would suggest sticking to Mint-that sleek orange published by Hindustan Times
3. Distinguish between learning and certifications:-Certificates are for the CV while learning is for your career. Stuff like NCFM/CFA is for the CV while mock trading/FRM is for the career, in terms of the learning value
4. The utility of NCFM:-NCFM modules are a good introduction to the financial markets, although they do depict an overly good-goody picture at times. Unless you need a certification for your work like some stock brokers/bolt operators do, the certificate itself is of doubtful utility even as a signalling mechanism
5. FLIP/other industry centered courses:-These often use multimedia, are current and practical oriented. Get feedback from other users/online forums, but these often are worth their price
6. Which qualification to pursue-CA/MBA/CFA..:-Assess yourself, match your abilities and objectives, also the time you have(can you take a full time break) and which degrees/schools help best for what. For example, FRM may be best for an aspiring risk manager, while a MBA from IIM Indore may be best for an Operations career and so on. This would need doing a lot of ground work, but no pain no pain.
7. Read Investor Letters/Transcripts/newsletters of experts:- Cisco's CEO John Chambers is renowned for his ability to 'call' the technology markets, Oaktree's Howard Mark's investor letters are the stuff of legend, and of course we have all heard of Warren Buffet's annual shareholder letters. But these are just the tip of the iceberg. In every field, be it self publishing or cooking, there are bound to be  experts who periodically share their expertise for free-even if on a time delayed basis. Use Google to find them, and then bookmark them.
8. Follow websites/blogs:-Thanks to their focus and independence, they can often give better analysis than any mainstream writer featured in the press.
9. Read financial history:-Books by Sorokin and a host of others, outline the history of crashes/booms/burst in great detail, and make the reader realize that this too will pass.
10. Don't forget finance is NOT the main driver: Even for banks/M&A advisory firms, offering finance is often just the first step/qualifier. It is technology, operations, marketing, strategy that makes the difference. Firms who have boosted their ROE with clever financial engineering(as a basic DuPont analysis would show!) have often left their shareholders holding the baby. Clever financial acumen is NOT a substitute for good business judgement. If you are a LBO fund, then maybe it is under the originate and breakup model, but not if you need to run the firm.
11. Get to know the faces/products behind the stock price; I cannot stress this enough! Many a time, a blue chip has stumbled because its leaders got arrested(Ansal/Everonn/DB Realty/LIC Housing Finance/Money Matters), management decided to spend on things like planes(Crompton Greaves) or merely because it was propped up by market manipulation(K10 stocks). And even the products/services matter. Because devoid of a clear market differentiators, that company will inevitably lose cash, and the management will have its ostrich head too deep in the sand to realize it, by which time it may be too late(like Money Matters).
12. Get a brokerage/demat account with Edelweiss:- I name Edelweiss-not because it is founded and owned by an IIM-A alumnus-but because it offers a great dashboard for seeing the fundamental parameters of a stock-and that too for free! Plus its data tools beat anything I've seen yet-and al that for free! Opening an account would get access to better functionality plus hopefully encourage them to keep those dashboards/tools free for others!
13. Identify your niche:- Some of us are great with words, others with numbers, yet others with that quick intuitive grasp of business or pattern recognition skills..the list can go on endlessly. For each of those skills, there is a great niche. For example. those good with words and a legal understanding, can pore through legal filings/notes to accounts to uncover value, as done by such sites(http://10qdetective.blogspot.com/). Those good with numbers can use DCF to good extent, those with business grasp can pick midcaps with growth potential/become VC investors, while those with pattern recognition skills can make their fortunes in technical analysis. All these are not only hobbies but great careers in themselves.
14. Do mock trading and maintain a trading diary:-In retrospect, we can rationalize to an amazing extent and even convince ourselves that we infact predicted events/did right things when we did not. Having a written record helps track those decisions, and base your faith in judgements on past empirical data. That said , past does not reflect the future as in those famous mutual fund disclaimers BUT note that even CAT/other objective exam preparation institutes suggest using past data on mocktest performance to refine test taking approach. So nothing wrong in it.